Professional Documents
Culture Documents
i
U
'
v,
z
u CASE F I L E
COPY
!y
t C. H. WodGPzg, Studey Faber, John J. Vun Buckel,
Cbades C. Olasky, Wayae K. Williums, J u h Le C. Mire,
and Jumes R. Homer
Munazed Spucecrafi Center
Hozcstoiz, Texas 77058 ? _d
The MSC Director waived the use of the International System of Units (SI) for this Apollo Experi-
ence Report, because, in his judgment, the use of SI Units would impair the usefulness of the
report o r result in excessive cost.
6. Abstract
Through space-flight experience and the development of simulators to meet the associated training
requirements, several factors have been established as fundamental for providing adequate flight
simulators for crew training. The development of flight simulators from Project Mercury througk
the Apollo 15 mission is described in this report. The functional uses, characteristics, and de-
velopment problems of the various simulators are discussed f o r the benefit of future programs.
Section page
SUMMARY ..................................... 1
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
CREW-STATION FIDELITY . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Controls and Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Stowage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Aural Cues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Markings and Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Space Suit and Cabin Environment Requirement . . . . . . . . . . . . . . . 13
Crew Couches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Crew-Station Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
VISUAL DISPLAY SIMULATION . . . . . . . . . . . . . . . . . . . . . . . . . 14
Display System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Celestial Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
FarBodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Target Vehicle f o r Rendezvous . . . . . . . . . . . . . . . . . . . . . . . . 19
Target Vehicle f o r Stationkeeping and Docking . . . . . . . . . . . . . . . . 20
Near-Body Scenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Landing Scenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
iii
Section Page
MOVING-BASE SIMULATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Simulation Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Configuration Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Configuration Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Configuration Accountability . . . . . . . . . . . . . . . . . . . . . . . . . 43
Concluding Re marks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
REFERENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
APPENDIX-SIMULATORDESCRIPTION ................... 47
iv
TABLES
Table Page
I FLIGHT CREW SIMULATORS. ..................... 3
FIGURES
Figure Page
Lunar module s i m u l a t o r . . . . . . . . . . . . . . . . . . . . . . . . . 5
14 Partial-gravity simulator . . . . . . . . . . . . . . . . . . . . . . . . 30
V
Figure Page
16 .......
Lunar landing r e s e a r c h facility (Langley Research Center) 31
vi
ACRONYMS
CM command module
EO engineering order
LM lunar module
,
t LMPS lunar module procedures simulator
I
vii
LMS lunar module simulator
MR modification request
TV television
3 -D three dimensional
viii
APOLLO EXPERIENCE REPORT
SIMULATION OF MANNED SPACE FLIGHT FOR C R W TRAINING
By C. H. Woodling, Stanley Faber, John J. Van Bockel,
Charles C. Olasky, Wayne K. Williams,
John L. C. Mire, and James R. Homer
Manned Spacecraft Center
SUMMARY
From the ear'ly phases of Project Mercury through the Gemini and Apollo Pro-
grams, flight simulators have been the key elements in the astronaut training programs.
As the missions progressed in complexity, the sophistication, number, and variety of
simulators employed f o r astronaut training were increased correspondingly. Through
space-flight experience and evolution of the simulators to meet associated training re-
quirements, several factors have been established as critical and basic f o r providing
adequate flight simulators f o r crew training. Included in these factors are high-fidelity
crew stations, especially in the area of controls and displays; accurate simulation of
the spacecraft systems, including the guidance computer and navigation system; com-
plete visual display systems for simulated out-the-window scenes; and certain moving-
base simulators for high-fidelity training in particular portions of the missions. The
significance of these factors f o r new programs will depend to a large degree on the mis-
sion objectives and requirements. Nevertheless, flight simulators incorporating some
of these items i n their design and operation will be vital in future astronaut training
programs.
INTRODUCTION
The topics of this report include some of the more significant aspects of space-
flight simulation: crew-station fidelity, visual display requirements, moving-base s i m -
ulations, and configuration management. The various topics relate primarily the
experience gained in these a r e a s of simulation since the beginning of manned space
flight. It is hoped that future programs might benefit through discussion of this
experience.
Credit for the separate sections of this report is given t o those who, through their
direct involvement with manned-space-flight simulation and training, were able t o re-
port firsthand the data contained herein. These individuals and their respective sections
are as follows: C. H. Woodling and John J. Van Bockel, background and discussion;
J a m e s R. Homer and John L. C. Mire, crew-station fidelity; Stanley Faber, visual
display requirements; Wayne K. Williams, moving-base simulations; Charles C. Olasky ,
simulator configuration management; and C. H. Woodling, overall compilation and
editing.
BACKGROUND
The crew-training simulators used f o r Project Mercury and f o r the Gemini and
Apollo Programs a r e listed in table I. These simulators a r e described in the appendix.
The crew-training use of the various simulators throughout the flight program is pre-
sented in tables 11 and III. During Project Mercury and the Gemini and Apollo P r o -
g r a m s , each crewman (on an average) spent one-third o r more of the total training
program t i m e in simulations (table 11). The crews of the lunar landing missions (Apollo
missions 11 to 15) spent slightly more than 50 percent of their total training time in
simulator training. In addition to the actual time spent in the various simulators, other
training activities were accomplished in support of the simulator training. For ex-
ample, the Apollo crews averaged more than 150 hours of systems briefings as a pre-
requisite to simulator training.
2
approximately 53 percent of the total simulation time of 1330 hours. In the Gemini
Program, the hours logged in the mission simulators are 67 percent of the total; and,
in the Apollo Program, the command module simulator (CMS)and lunar module simu-
lator (LMS) hours combined are 80 percent of the total of 29 967 hours spent in all
Apollo simulations.
Centrifuge (NASA Manned Spacecraft Center) Moving-base simulator f o r launch, laun:h abort,
a d em with associated acceleration
envimnment
I*my larding trrirri vehicle (LLTV) Free-flight trpining vehicle for final phase of lunu
Mint?
Lunar larding training vehicle simulatw W m i l i a r i u t i m with LLTV myatems
hagley lunar lading research facility Dynamic simulator (tetbered) f o r final phpac d
l W hnding
Dynamic simulator f o r W i n g in tk V6-g lunar-
surface environment
3
TABLE n.- SIMULATOR TRAINING SUMMARY
Gemini 17 991
1 Simulator
~~
Time per program, hr
~
Apollo
Total
t i n e , hr
Mercury Gemini (through
mission 15)
Mercury procedures simulator (2) -- 708
Centrduge 58 496
Part-task trainer -- 61
aTime computed on the basis of 2 hours logged for each LLTV flight.
bNot included in total simulator hours.
4
C I assi fi cation
Simulators can be grouped into several classifications. For the purpose of this
report, the following definitions will apply.
simulator (GMS) (fig. 2), the CMS (fig. 3), and the L M S (fig. 4).
5
Part-task simulator. - A part-task simulator is a device in which only a portion
of the mission o r several major crew t a s k s are simulated. Fidelity in these t a s k s may
be greater than that achieved with the mission simulator.
D i scussion
The value of high-fidelity simulation
A
L was well known through aircraft flight ex-
perience before Project Mercury. The de-
pendence on simulation for space mission
success and crew safety generally is g r e a t e r
than the dependence on simulation for air-
craft testing because of the natures of the
two flight programs. Space-flight crews
are fully committed at lift-off to an entire
mission, in which a broad envelope of opera-
tional variables is exercised. Aircraft t e s t
r e s e a r c h usually allows for a more gradual
exercising of the total flight envelope
through a series of "buildup" flights.
Whereas the a i r c r a f t test pilot can obtain
much of his training coincident with actual
Figure 5. - Lunar landing training flight testing, the crews f o r space missions
vehicle. must receive all their training and be highly
proficient in all flight tasks before the mis-
sion. Primarily f o r this reason, the space-
flight simulators require the highest degree of fidelity of spacecraft and mission
simulation. Aircraft experience was used extensively in the development and operation
of the first space-flight simulator, the MPS. For the follow-on space programs,
Gemini and Apollo, the major developmental impetus f o r the simulators was derived
from space program experience,
Early in Project Mercury, the effect of space -flight environmental factors upon
crew performance caused considerable concern. Consequently, training emphasized
6
crew exposure t o such conditions as high acceleration forces, zero-g conditions, heat,
noise, and spacecraft tumbling motion. The Mercury astronauts received many train-
ing exercises in an attempt to duplicate these environmental conditions. Particular
concern was expressed about crew capability t o manually control the spacecraft during
the high acceleration loads imposed during launch and entry. As a result, the Mercury
astronauts participated in four centrifuge programs at the Naval Air Development Cen-
ter, Johnsville, Pennsylvania. Project Mercury flight results verified that the condi-
tions of space flight had no adverse effects on crew performance for missions as long
as 22 hours in duration. Consequently, crew-training programs for the Gemini and
Apollo missions deemphasized such considerations and concentrated t o a great extent
on the many and complex operational considerations. Project Mercury experience did
point out the importance of simulating out-the-window scenes, which became even more
important f o r the Gemini and Apollo missions. The Mercury simulators did not have
adequate out-the-window displays, and a major effort to remedy this situation f o r the
Gemini simulators was begun.
The Gemini Program required more sophisticated simulators than did Project
Mercury, principally because of the Gemini rendezvous mission objective. Crew capa-
bility to effect the rendezvous by using primarily out-the-window information was essen-
tial to the mission and dictated that an elaborate visual system be incorporated into the
Gemini mission simulators. The development and use of a n advanced state-of-the-art
infinity optical display system added considerably to the realism and value of simulator
training for the Gemini crews.
Results of the Gemini Program indicated quite clearly that a well-defined and ef-
fective configuration management system was needed t o maintain the simulators in a
configuration that corresponded closely with the continuously changing spacecraft.
Basically, this meant a quick-response systemoperated by personnel cognizant of all
spacecraft changes and capable of deciding on and contracting f o r the necessary modi-
fications and incorporating these into the various simulators. A configuration control
panel committee, with ancillary working groups, was formed for the Gemini and Apollo
P r o g r a m s with good results.
7
TABLE IV. - SIMULATED NETWORK SIMULATIONS~
~~ ~
I11 4
IV 7
V 7
VI 13
VII/VI 9
VI11 9
Ix - 9
X 7
XI 6
XI1 7
7 18 0 18
8 14 0 14
9 10 8 20
10 11 7 18
11 7 7 18
12 10 12 25
13 13 9 27
14 15 13 35
15 19 7 31
a
This summary includes only the mission simulations involving the use of mis-
sion simulators with flight crews and does not include the many more additional days
of "math model" simulations (without the mission simulators) for flight controller
training.
Forecasts of the simulator training requirements for the Apollo Program showed
that a large inventory of various types (full mission, part task, moving base) of stra-
tegically located simulators would be needed. The Apollo Program imposed severe
schedule and simulator-complexity constraints; a requirement was dictated to train a
large number of three-man crews within a relatively short time in sophisticated and
8
critical mission simulators (three command module simulators and two lunar module
simulators) with complete visual systems. The capability for integral operation of the
simulators with each other and with the MCC in Houston a l s o was required. The greater
number of command module simulators was needed to accommodate the greater number
of command-module-only flights that comprised e a r l y program concepts. In addition,
certain special-purpose part-task simulators were considered necessary to provide
supplementary training, to afford overall training flexibility, and to support the crew
procedures development effort (an integral part of crew training). Among these special-
purpose simulators were the command module procedures simulator (CMPS) (fig. 6),
the lunar module procedures simulator (LMPS) (fig. 71, the Apollo translation and dock-
ing simulator (TDS)(fig. 8), and the Apollo dynamic crew procedures simulator (DCPS)
(fig. 9). Perhaps the most notable special-purpose simulator was the LLTV. Because
of the limitations of accurately simulating this critical phase of the lunar mission on
ground-based simulators, the free-flying LLTV was employed to provide an extremely
realistic dynamic simulation of the lunar approach and touchdown.
Concerning the simulators themselves, s e v e r a l basic decisions were made t o
satisfy the fidelity requirements. One of these decisions was that all spacecraft suh-
s y s t e m s would be simulated by the com-
puters; that is, no actual (hardware) space-
, craft subsystems would be used. One area
handled in a special way was simulation of
the Apollo onboard guidance and navigation
computer. The simulation successfully
used was based on an "interpreter" concept
9
Figure 8. - Translation and docking
simulator (Apollo) .
in which a general-purpose digital com-
puter was programed to accept the same
program as the flight computer and to re-
spond to spacecraft systems precisely as
the flight computer would. The interpre-
ter superseded a barely adequate, costly,
and always late functional approach for Figure 9. - Dynamic crew procedures
simulation of the onboard computer. simulator (Apollo) .
Another area that required considerable updating and redesign was the difficult
visual simulation of the lunar landing. As will be discussed in the section on visual
display simulation, the initial concept of the LMS fell short of the realism required.
An intensive development effort was undertaken to improve the visual simulation for the
lunar landing and at the same time to permit timely use of the LMS f o r the early orbital
training. This effort was successfully concluded t o provide high-fidelity landing and
ascent training for the first lunar landing and subsequent flights.
Although the remainder of this report is oriented toward the difficulties and ex-
perience gained through the operation of the simulator hardware, it by no means implies
a lesser importance of the software systems. Indeed, a large percentage of the overall
simulation effort throughout all flight programs was associated with the simulation soft-
ware. For example, the Apollo lunar landing mission simulation for the CMS consisted
of 750 000 words of memory residing in four large digital computers. Simiiarly, the
LM simulation in the LMS required 600 000 words in three digital computers. At the
peak of preparation f o r the first lunar landing mission, approximately 175 support con-
t r a c t o r personnel were assigned to the development and control of the software systems
of three command module simulators and two lunar module simulators. Another
200 contractor personnel were assigned to the hardware operations and maintenance.
Simulator changes required t o keep pace with the changing spacecraft and mission num-
bered more than 30 p e r month in this same time f r a m e . Configuration management of
these hardware and software s y s t e m s is discussed in the section entitled "Simulator
Configuration Management. 'I
10
CREW-STATION FIDELITY
Various considerations for the design of the simulator crew station are described
in this section. The Mercury, Gemini, and Apollo simulator crew stations, particu-
larly those for the mission simulators, were designed and constructed with a high de-
g r e e of fidelity to simulate closely both spacecraft performance and interior appearance.
11
Stowage
Since the early Mercury flights, the proper location of onboard data and loose
equipment has been a necessary inflight task and, therefore, a training requirement.
The need for assigning every item a specific stowage location and the procedure for
handling loose items could not be neglected. Ideally, flight-type equipment f o r stowage
training is preferred. However, because of the disadvantages of the high cost of the
equipment and the frequent handling under one-g conditions, actual flight equipment is
seldom used for simulator stowage. For training purposes, a nonflight item s e r v e s as
well if it functionally fulfills the intended use.
Lighting
Lighting of the crew compartment was not considered critical in t e r m s of intensi-
ties and spectrum composition. Experience has shown that light levels sufficient for
general illumination are satisfactory as long as the simulated intensity does not deviate
appreciably (+20 percent) from the actual.
Because of the use of optical eyepieces (telescope and sextant) for navigational
sighting in the Apollo spacecraft, the floodlight intensity settings were usually low,
which caused some difficulty in reading the meters. As a consequence, all simulator
I m e t e r s were lighted integrally, and panel nomenclature was blacklighted with electro-
luminescent elements, as in the spacecraft.
Aural Cues
Audible cues are simulated and presented to the flight crews whenever applicable
during the training exercise. On the Apollo mission simulators (CMS and LMS), such
cues as booster thrust, cabin decompression, and the firing of pyrotechnic devices and
~
attitude-control jets are simulated. Noise levels in the communication channels as well
as ambient noise are simulated and transmitted to the flight crews through headsets and
I
through concealed loudspeakers within the crew stations.
12
Space Suit and Cabin Environment Requirement
In the various simulators, suited operation throughout all phases of flight, yet
with the capability for the crew to train unsuited for comfort, has been emphasized.
The crew- station environment is maintained at ambient p r e s s u r e with circulated,
temperature-regulated air. All the controls f o r the crew-station environment are man-
ual, and the systems are purely simulator oriented. The crew-station-environment
parameters of temperature, humidity, and noise level are electrically monitored at the
instructor-operator station. The crew-station cabin-temperature and cabin-pressure
m e t e r s display simulated values in accordance with the flight profile; that i s , while on
the launch pad, the simulated cabin pressure is 14.5 psi and decreases proportionately
during launch, settling at 5.0 psi for orbit conditions.
Crew Couches
Exact replicas of the crew couches have been used in the simulators to provide
proper body restraint, correct reach patterns, and accurate couch-manipulation
capability.
Crew-Station Hardware
Simulator crew-station hardware is discussed in two categories : contractor-
furnished equipment (CFE) and Government-furnished equipment (GFE).
In many cases, simulator parts do not require the sophisticated design o r the
rigid tolerances demanded for the spacecraft. The frequent use of rejected spacecraft
p a r t s on the simulators results in cost savings. During the Apollo Program, the space-
craft contractor maintained another manufacturing capability commonly referred to as
the "model" shop. The personnel in t h i s shop usually work from engineering design
sketches to fabricate prototype units f o r spacecraft qualification tests. The model shop
is often used to fabricate simulator hardware. Components made in this shop f o r the
13
simulators do not have to meet the stringent spacecraft specifications. Occasionally,
hardware f o r the simulators is made from initial design sketches to meet the desired
leadtime necessary to support crew-training exercises. Various other contractor
.sources of simulator hardware are available, including the simulator manufacturer.
Concluding Remarks
A high degree of crew-station fidelity, particularly in the full-mission simulators,
is necessary for accurate simulation of spacecraft performance and mission character-
istics. In addition to a complete simulation of the spacecraft controls and displays,
such items as stowage provisions, lighting, aural cues, and crew hardware are ex-
tremely important in providing this required fidelity.
Early in the evolution of manned space flight, the requirement f o r a window in the
spacecraft w a s identified. Although the window served several purposes, the most im-
portant was as a backup attitude reference. Experience in aircraft control t a s k s and
during early space flight had established that training was necessary t o enable the crew-
member to correlate his out-the-window references with his instruments. The various
techniques used to produce the simulated out-the-window scenes throughout Project
Mercury and the Gemini and Apollo Programs are described in this section.
Display System
The first element of a visual simulation system is the display system, the element
into which an image is projected for viewing by the crew. The determining factors of
this element are the apparent distance t o the image being viewed and the field of view
(FOV) of the window. A general requirement is that all images appear at infinity as
they do in reality. For docking, the simulation requirement was that the image appear
t o approach the crewmember as in the actual situation. Field-of -view requirements
were based on the particular spacecraft window being simulated, with some allowance
f o r pilot eye o r head movement. The Mercury spacecraft requirement was a 60" FOV,
although that of the Apollo LM (the largest requirement) was a 110" FOV as measured
at the eye.
14
All mission simulators used an infinity display system. Figure 10 is a schematic
of a typical window system using reflective optics techniques. To satisfy the docking
display requirement, the display or output tube was moved from its infinity position.
Approximately 4 inches of movement of the display tube made the image appear to move
f r o m infinity to 8 feet from the viewer.
The infinity image system has the
desirable characteristic that, as the crew-
man moves his head, the item being viewed
appears to move correctly. Conversely,
because of parallax, viewer head motion
grossly affects what is seen with direct-
view, screen-type systems; therefore, the
viewer must hold his head relatively still
the source of the ob-
In the LMS, the large FOV, coupled with a large window (17 by 22 inches to ex-
t r e m e apexes) and almost total freedom of the crewman to position his head, required
the greatest compromise. The first compromise was to place the exit pupil in the area
of the crewman's design eyepoint (i. e. , the normal eye position). From this design
eyepoint, the crewman could move approximately 4 inches in any direction with no loss
of display. The second compromise was to point the axis of the forward-window display
down to feature the landing scene at the expense of the upward scene. In the actual
spacecraft, the crewman can look up approximately 20" from the design eyepoint, and
even more by hunching down. In the simulator, the crewman was limited to less than
10". Even with this compromise, the bottom apex of the LM window was blank as
viewed from forward eyepoints. Because these adjustments were chosen very carefully
against anticipated use of the window, the limitations have had no adverse effect on
training.
15
The infinity display systems used with the mission simulators produced the de-
sired effects as long as proper consideration was given to mission simulation require-
ments. These systems did have some drawbacks. Relatively high cost (30 t o 40 percent
of total mission simulator cost), large physical sizes, ghost images, extremely low
light transmission, and backlight scatter were the most significant. Since the initial
application of the infinity optics technique to simulation, additional development has re-
sulted in reductions of cost and size. These new systems are of both refractive and
reflective types. The refractive types tend t o be brighter, although with some loss in
image quality, than the reflective types.
Celestial Simulation
Simulation of the stars was predicted on both navigational and attitude-reference
requirements. In orbital flight, these requirements are intermixed because the pilot
must know h i s location to use the stars as an attitude reference. In simulation of the
star field, consideration was given to the number and relative brightness of the stars
selected and t o the static and dynamic accuracies required. The number of stars chosen
for all projects to date is approximately 1000, consisting of all those brighter than a
+5 magnitude. The logic of selecting this number was that the navigation stars and the
constellations used for identifying them are all brighter than the fifth magnitude. One
other factor in this selection was that fifth magnitude stars are the dimmest generally
visible t o the unaided eye from the ground.
Static positional accuracy specifications were dictated by the need t o identify con-
stellations. It w a s determined that the appearance of constellations changes signifi-
cantly with even minor positional e r r o r s . Furthermore, if the s t a r s were to be used
in a navigational task, an instrument such as a sextant would be necessary and an accu-
racy compatible with the instrument was required. To reduce cost, star positional re-
quirements were moderated to 1 milliradian (approximately 0.06") f o r all navigation
stars and 0.5" for all other stars. Specifications also required that if the constellation
appeared incorrect with a 0.5" e r r o r in a given star location, the positional accuracy
f o r that star must be tightened to that of the navigation star.
The specified dynamic positional accuracy also required compromise, as the re-
quirement was complicated by the large range of vehicle angular motions possible in a
spacecraft. Pilot-controlled rates range up t o 20 deg/sec, with uncontrolled vehicle
angular rates of 50 t o 60 deg/sec possible. Although a perfectly stationary spacecraft
is not possible, rates can be as low as 0.1 deg/min f o r small spacecraft motions.
From h i s stable and vibrationless viewing position, the crewman could detect such small
motion. The final specifications derived for all spacecraft simulation programs re-
quired as smooth a low-speed drive as could be produced and permitted lags in position
at angular rates above 1 5 deg/sec. Stepping motions that were detectable by the crew-
members were deemed unacceptable.
16
In all simulators, the same basic technique was used to produce the star displays.
Approximately 1000 individual stars were simulated by small, highly reflective steel
balls set into the surface of a (celestial) sphere. When the highly polished balls were
illuminated by a point source of light, they reflected the point of light, the relative
brightness being a function of the size and coating of the balls. Balls as small as
0.1 inch and as large as 0.8 inch in diameter have been used. Mirror surfaces concen-
t r i c with the celestial sphere were used t o keep the individual balls in focus over the
FOV of the window. This technique produced excellent representations of the star
field. The only anomaly noted on the display was a halo created by some of the coated
balls. Unfortunately, in the LMS the halo stars were all navigation stars, and the halo
made recognition relatively easy. Therefore, the use of coatings to simulate various
magnitudes is not recommended.
A chronic difficulty in the simulation of stars was in the dynamic drives of the
celestial spheres. Smooth, stepless servomechanism drives that hold positional cali-
brations have not been obtainable, and only continuous maintenance and adjustment have
produced acceptable operation. This problem has not been limited to the servos and
torque motors but has been noted in the entire chain, from the equations of motion,
through the digital conversion equipment, to the spheres. The problem has become
more acute with each new program, because the range of dynamic motion produced by
crew action and under crew control has increased from approximately 10 deg/sec for
the Mercury spacecraft to approximately 20 deg/sec f o r the LM. Although, in the past,
continual maintenance and frequent calibration have tended to minimize the impact of
the servomechanism problems, additional research in product improvement as well as
improved mathematical and computational techniques should be continued.
In addition to the window and the unity-power telescope, the command module (CM)
required simulation of a 28-power, 2" FOV sextant. This optical instrument directly
interfaced with the onboard guidance computer f o r navigational procedures. The r e -
quirement f o r the sextant simulation w a s to produce the Apollo navigation stars (56 in
number) with c o r r e c t background stars, as well as to provide earth o r moon limb at
various altitudes. Static accuracies were the same as specified f o r the spacecraft,
that i s , 10 arc seconds. The dynamic accuracy requirements of the sextant simulation
were not high, because all tracking and marking t a s k s were performed at very low scene
dynamic rates. The sextant simulation was produced by using a single extremely accu-
rately positioned navigation star in one optical leg and a selection of as many as
90 slides of star fields, limbs, and landmarks in a second leg. A pattern generator
and filters were used to produce the background fields f o r the navigation star. Pairs
of rhombic p r i s m s were used t o produce the motions in both legs. The two legs were
combined into a single image by a beamsplitter assembly. For most Apollo training,
five slides covering the most-used navigation s t a r s were used in the second leg. In
operation, if either leg were more than 1" off the intended line of sight, the sextant
simulation was turned off.
17
was not used. As a result, it was not possible for the crewmember t o identify the par-
ticular navigation star in the sextant. Fortunately, this turned out to be only a minor
limitation. The accuracy and stability of the actual spacecraft guidance, navigation,
and stabilization systems were quite good, and the inflight procedures were developed
t o make most effective use of this spacecraft capability. That is, the spacecraft crew-
member was not required to identify a navigation star with the sextant, because his on-
board guidance computer programs ensured that the navigation star was always within
0.5 of predicted value.
O
Far Bodies
F o r the purpose of this discussion, far bodies are identified as the distant moon,
earth, sun, and planets, that i s , all those bodies in the s o l a r system that subtend an
angle of approximately 0.5" from the viewpoint. These far bodies were simulated for
navigational reference and for added realism. In the Apollo simulation program, they
were deemed important enough t o be included. F o r the sun, the specifications were f o r
position, size, and high relative brightness in the windows when the individual window
was pointed toward the sun. The requirements for distant e a r t h and moon were f o r po-
sition, brightness compatible with star simulation, and apparent size change as a func-
tion of range t o each body. In addition, simulation of the sun terminator on the far body
was desirable. The tolerances on these requirements were not tight. Simulation of the
larger planets, although desirable f o r navigational reference, was not mandatory.
As noted previously, these effects were not included in the Mercury and Gemini
simulators. In the Gemini simulator, the sun in the window o r , r a t h e r , the total wash-
out of the scene caused by the sun in the window was simulated by video flooding the
television (TV) cathode-ray tube (CRT).
The sun simulation in the CMS used a r c lamps to produce a 0.5" spot of light in
each of the windows. Although these lamps did not duplicate the exact sun brightness,
the simulation was sufficient to wash out the stars and other dim images and served to
remind the crew to maneuver t o some other attitude if they desired t o perform any out-
the -window visual tasks.
With the LMS, a high level of sun brightness was not required, and the sun was
simulated by a s m a l l disk attached to the celestial sphere. By this means, the desired
effect was provided at much s m a l l e r computation and hardware cost than in the CMS.
Sun lighting (or sun shafting) in the crew station was not provided in either the CMS o r
the LMS. Although this omission in the simulation has resulted in the crews' using
spacecraft lights in the simulators that are not normally required in flight, the problem
has not been a major one.
The simulation of the distant moon and earth in the Apollo mission simulators also
fell short of the original specifications. In the CMS, a direct-view film system with
28 discrete phases of the distant moon w a s defined. Because these images were too
small t o be reproduced on film, the technique has been abandoned. For most CM simu-
lation work, the sun simulator was used t o represent the distant body. In the LMS, a
technique similar to the LMS sun simulation was used, but the disk was somewhat less
reflective than the solar disk. By shaping the disk, the phases of the moon o r e a r t h
also were presented. The distant e a r t h and distant moon were moved o r manually
18
adjusted to provide for movement of the distant body relative to the stars, changes in
sun terminator on the body, and changes in relative size with variation in distance.
In summary, the far-body simulation requirements have been modified to be con-
sistent with crew-training needs and state-of -the-& hardware. The requirements for
far-body simulation from the eyepoint are summarized as follows:
5. Dynamics - Far body should have no apparent motion relative to the star for
spacecraft-attitude motion simulation. Far body should move with respect t o the stars
for long-term emphemeris effects; however, step changes are acceptable.
6. Washout and sun shafting -Washout and sun shafting are desirable.
The target vehicle f o r distant rendezvous does not require attitude information.
The two requirements are position in the window and relative brightness with range.
F o r tracking in darkness, the spot of light was required to blink to simulate a flashing
beacon. For daylight tracking, the spot should be steady t o simulate reflected sunlight.
Other factors a r e an accuracy of 4 " on position, minimum brightness equivalent to a
fifth-magnitude star, and a maximum range of 250 miles for the unaided eye. Maximum
brightness was not specified; however, the intent was to make the minimum-range tar-
get as bright as possible. The spot of light should be small enough that it does not
stand out against the background stars unrealistically.
In all simulators with rendezvous capability, the window simulation was produced
by electronic techniques on a CRT. This technique met the requirements. The major
problem was in positional stability of the display. In the early simulations, frequent
alinements were required; however, over the span of the Gemini and Apollo Programs,
improved electronic circuit design has reduced this problem t o a minimum.
19
In the Apollo CM, the unity-power telescope and a 28-power sextant were used
during rendezvous. The simulation requirements for these instruments are similar
to that of the basic window except f o r positional accuracy. The telescope was used to
point the sextant t o a specified accuracy of 0.5" relative to the sextant line of sight.
For compatibility with navigational requirements, the sextant accuracy was specified
as 30 arc seconds.
The telescope simulation f o r distant bodies was a later addition to the original
simulator. Because of space constraints and the stability problem discussed previously,
a CRT display w a s not practical. Instead, the point of light representing the target ve-
hicle was produced by a simple light bulb. Displacements in the FOV and size were
produced by mechanical means.
The CM sextant simulation also was added to the original simulator design. The
requirement of the distant target was satisfied by using the sextant navigation star s i m -
ulation (described in the paragraphs on celestial simulation) t o represent the target-
vehicle position. The resultant positional accuracy was much better than 30 arc
seconds. There were two limitations with this sextant simulation. First, with
28-power magnification, the size, shape, and attitude of the target vehicle were appar-
ent f o r relatively long ranges. Second, the omission behind the LM of a lunar back-
ground made spotting the LM much easier in the simulator than in flight. Of the two
limitations, the lack of a lunar background was considered more critical. To fully
simulate the actual situation, a moving scene beneath the CM and behind the LM would
be required. However, a full simulation is not mandatory. A static scene of the lunar
disk, o r of the lunar limb and terminator, fulfills the minimum requirements.
Two general techniques have been used for target-vehicle simuiation: a direct
analog system of closed-circuit T V and models, and an electronically generated (drawn)
image. In both systems, the input to the display system was through a CRT in the in-
finity optics systems. The electronic image generator (EIG) was used successfully in
one of the Gemini mission simulators, in the Gemini part-task t r a i n e r , and in the
LMPS. In the EIG system, the target vehicle was drawn on the face of the CRT. The
outline o r envelope of the target was drawn at a 60-hertz rate; however, the surface
was filled in at a 15.75-kilohertz rate. The image generation contained nine degrees
of freedom and produced such phenomena as line -of -sight blanking, illumination
20
shadowing, and perspective distortion. Simple target shapes (cylinders, cones, and
others) as well as combinations of these shapes were readily simulated with simple
surface markings and details.
In the more conventional system such as that used on the Apollo mission simula-
t o r s , a scale model of the target was viewed by means of a closed-circuit TV system.
The model was mounted in a three-gimbal system to reproduce target-vehicle rotations.
Generally, the innermost gimbals are in the model. The gimbal system should be off-
set relative t o the c a m e r a optical axis to avoid gimbal lock at docking attitude. The TV
c a m e r a s o r lens s y s t e m s attached t o the TV cameras also are gimbaled in this c a s e t o
represent the active-vehicle rotations. In the CMS, however, r a t h e r than c a m e r a rota-
tion, active-vehicle motion was introduced by displacement of the video raster on the
display CRT.
The relative range was introduced through a combination of techniques. In all TV
systems, a t r a c k and carriage assembly was used to move the c a m e r a away from the
model. Camera motion produced an apparent model-size reduction. For the Apollo
CMS and the Gemini mission simulators, this size reduction was supplemented by a
raster-shrink technique t o provide total changes in range. Raster shrink of 67 t o 1 was
successfully used in the CMS. The L M S used two models, a one-eightieth and a
one-twentieth scale. The total required range or apparent size change was obtained by
moving away from one model, then switching to the other and moving from it. An ex-
ample of the type of model used is shown in figure 11. This model, which is from the
LMS, contains the lighted docking t a r g e t in one of the CM docking windows. The dock-
ing probe and the CM conical portion were made of rubber to protect the probe and
c a m e r a s in the event of inadvertent collision.
21
.
With the TV/model technique, external features such as windows, antennas, and dock-
ing aids can be portrayed very accurately. These features are extremely important,
especially at the close-in docking distances. The picture fidelity obtainable with the
TV/model system must be weighed against the complexity of the electromechanical de-
vices required by this system. The TV/model technique has been animated s o success-
fully in the mission simulators that full-scale-vehicle simulation f o r docking is no
longer considered necessary, as it was in early Apollo training.
Near-Body Scenes
Scenes of the near body while in orbital flight cover the altitude range above the
e a r t h o r the moon from 8 to approximately 1000 nautical miles. At these altitudes, the
surface was assumed to be reasonably smooth, and t e r r a i n that appeared three dimen-
sional (3-D)was not required. Near-body s c e n e s were necessary f o r two purposes.
The first was as a general reference from which the pilot could evaluate his attitude
and estimate his position over the t e r r a i n . To meet these needs, horizon position,
groundtrack movement, and general continental features were required. The second
purpose was f o r landmark identification and tracking. To meet these requirements, a
more detailed terrain scene was necessary, and the spacecraft position relative t o the
scene had to be more exact. For the earth, color was a highly desirable feature in ter-
rain identification. For either function, several additional features were desirable,
such as a day -night terminator and the highlight brightness that would approximate the
out-the-window scene. Regarding the accuracies, the following p a r a m e t e r s are noted :
Scene
a
a
+-j-Y
Attitude function
10
0.5
--
I
Landmark function
.5
.1
Landmark area was defined generally as a 10-mile-radius circle about the spe-
cific landmark. The transition from the accuracy of the landmark area to the general
scene should occur over a 100-mile radius.
22
For the attitude function, the allowable distortion should permit identification of gross
continental features. F o r the landmark function, the allowable distortion should permit
easy identification of landmass features, capes, bays, peninsulas, r i v e r basins, major
craters on the moon, and s o forth.
In Project Mercury, the view through the periscope and through the window was
simulated in what might be termed "moving-base devices." The periscope was simu-
lated i n the air-lubricated free-attitude (ALFA) trainer (fig. 12) by viewing a 12-foot-
diameter rear-projection screen. The nadir scene was projected on the screen from
a hand-painted strip film. Considering the large distortion of the spacecraft and t r a i n e r
periscopes, the basic requirements of attitude reference and recognition of g r o s s con-
tinental f e a t u r e s were met. The requirement for window simulation in Project Mercury
was uniquely tied to yaw-angle identification while i n retrofire attitude. This need arose
f r o m an inflight problem during one of the e a r l y Mercury flights. Before the following
flight, a yaw recognition device had been conceived and fabricated. The simulation con-
sisted of a 32-foot-diameter s c r e e n curved t o represent a portion of a globe 63 feet in
radius. The shape was obtained by inflating an airtight envelope consisting of one
translucent and one transparent Mylar sheet. A filmstrip depicting moving cloud cover
was projected onto the translucent screen. The crewmember standing at the center
and approximately 2 feet from the dome was at the proper scaled altitude. A simple,
hand-held box outlined his window. Using this simulation, the crewmember could ob-
serve any yaw angle and readily learn to
identify the yaw angles of interest while
looking toward the horizon.
23
positional accuracy and curvature up to a n altitude of 4000 nautical miles. An artifact
in the globe and probe system was the need t o nutate the probe in azimuth as a function
of latitude, an effect that complicated the probe drive equations.
The LMPS used a flying spot scanner system (from the KSC-based GMS) with film
as the data-storage element. Latitude and longitude were produced by film translation,
and spacecraft motions were introduced by electronic manipulation of the flying spot
scanner output. The scanner pattern w a s s p i r a l and centered at the nadir to produce
a clean, sharp horizon. To provide equal element spacing as measured from the view-
point in the spacecraft, the spiral scan spacing was nonlinear. The design resolution
w a s equivalent to the best high-resolution TV raster systems. In the GMS, two parallel
systems were used to produce a semblance of color. However, in the modification to
adapt this system t o the LMPS, this provision was eliminated.’ The modification also
allowed simulation of an altitude range f r o m 2500 feet to 100 miles by using zoom optics,
zoom electronics, and a series of various film scales. The remainder of the LMPS
modification was directed toward improving overall quality of the visual display by using
a l a r g e r f i l m format and improved electronics. Even with these improvements, the
LMPS system has not proved as stable and as flexible as the more conventional mechan-
ical and optical systems.
The LMS used a system s i m i l a r in certain aspects to both of the previously de-
scribed systems. In the LMS, a n articulated probe and closed-circuit T V viewed a
filmstrip of the near body. The LMS films and the LMPS films were actually the s a m e
image material printed at slightly different scales. The major difference in the s y s -
t e m s was that, in the LMS, four windows were active and four probes were required,
whereas, in the LMPS, the display was provided to one forward window and to the over-
head window. Furthermore, the LMS scene requirements included landmark identifica-
tion and tracking in addition to the vehicle attitude-reference task. In the LMS, a
common filmstrip was projected through zoom optics onto four screens. The altitude
range from near z e r o t o orbit was simulated by a combination of optical zoom and film
changes. The probes were mounted at a fixed distance above the s c r e e n s and articu-
lated with the ability to scan to any point on the screen. At the higher altitudes, spher-
ical distortion was introduced by moving additional optical elements into the projection
chain. The horizon was produced at the higher altitudes by illuminating an area smaller
than full-screen size. At altitudes below 100 000 feet, a servomechanism-operated
m i r r o r system surrounding the screen produced the horizon. The lunar scene films
used were, in chronological order, artistic rendition of the front side coupled with
artistic imagination for rendition of the back, artistic rendition updated with Lunar
Orbiter data, Lunar Orbiter photographs photomosaicked into a filmstrip, and Lunar
Orbiter photographs photomosaicked with lunar photographs from Apollo flights. Final
configuration of the film consisted of Lunar Orbiter s t r i p s f o r the high-altitude, full-
orbit scenes and Lunar Orbiter s t r i p s mosaicked with Apollo photographs for the low-
altitude scenes in the vicinity of the lunar landing site.
Several problems were associated with the system that materially reduced the
usefulness of the LMS. A problem resulted from the splitting of the film image to four
screens. The illumination on each vidicon was one-third t o one-half of the desired min-
imum. Another difficulty involved the zoom optics. The design required no movement
of the optical axis with zoom. The dynamic wander of the original hardware was such
that the usable zoom ratio was restricted t o 3 from the design ratio of 10. The wander
a l s o made registration difficult t o maintain when switching from one to a second film
24
scale. A problem was also associated with the film graphics. It was not practical to
obtain continuous filmstrips with the information content desired. The Lunar Orbiter
s t r i p s with Apollo photographs were available only f o r small areas of the moon. Also,
framelet lines in the Lunar Orbiter s t r i p s were reproduced just as vividly as the craters
and rilles. Furthermore, in the Lunar Orbiter films, it was not possible to simulate
the effect of the sun elevation angle. Hand-painted films could have alleviated certain
of these problems; however, the cost in manpower and time was prohibitive.
The overall effect of these problems w a s a degradation in the LMS fidelity, which
necessitated use of other techniques and facilities to complement the LMS. F o r in-
stance, f o r the landmark identification and tracking task, detailed briefings and lunar
map reviews were held with the crew. The accuracy of the spacecraft navigation sys-
I tem also helped, because the crew rarely had t o search for a landmark during flighty.
I The desired target was usually where mission planning said it should be.
c
The CMS near-body simulation, from the outset, was mainly for the purpose of
: landmark tracking. The system consisted of direct view of color film projections of
near-body imagery. Only such a direct-view system could provide the resolution spec-
ified. This system, known as the mission effects projector (MEP), w a s used in the
four cabin windows and in the unity-power telescope. In the MEP, the image was pro-
jected through a series of optical devices onto a spherical screen. This screen served
as the input of the infinity image system. The optics were driven by servomechanisms
I to simulate spacecraft rotations, limited altitude effects, day-night terminator, and so
forth. Nadir position was introduced by driving the film, and large altitude changes
were provided by dissolving t o films with different scale factors. The MEP represented
an extremely complex electronic, mechanical, and optical device. For instance, each
,
I
MEP had over 30 computer-controlled servomotors and over 40 other computer-
generated signals. Some compromises were required in the final operating configura-
tion: the maximum detail scenic area, as measured from the spacecraft, was limited
to a 90" cone centered about the nadir; the minimum altitude simulated was 100 nautical
1 miles: and the film centerline was placed approximately along the groundtrack. The
first of these was the greatest limitation, but it was partially alleviated by using a cloud
1 cover that moved with the earth to f i l l the scene to the horizon.
The films covered relatively narrow strips centered on the groundtrack; and, be-
cause of retrogression of the orbit (westerly movement of the ascending node), 17 con-
tinuous s t r i p s were required to cover the Apollo earth-orbit mission. To resolve the
round-to-flat mapping problem and the orbit-to-orbit scene repeatability, an extremely
, accurate globe, used as the data source, was photographed by a special modified-slit
camera. The preparation of the filmstrip from this globe is described in reference 1.
The altitude variation for near-body orbits was specified as 100 to 1000 nautical
miles and was obtained by a combination of three film scales with a 2.15:l zoom capa-
bility. Over 250 feet of film stored in four film cassettes were required f o r each MEP
system.
The CMS MEP system, like the LMS, did not meet all the initial design objectives.
The complexity of the mechanical-optical device caused problems, and the number of
MEP systems required to f i l l the windows of the three mission simulators made i m -
provements extremely costly. The first of the difficulties concerned the illumination
25
system, in t e r m s of both the level of light and the uniformity of illumination a c r o s s the
full FOV. The relatively low level of illumination manifested itself in many other prob-
lem areas, The system design required an extremely high luminous f l u on the surface
of the f i l m . Even with this flux, the final brightness as measured from the window was
approximately one-fourth of the originally specified 4 . 5 ft-L and approximately
one -eighth of the desired illumination t o simulate a highlight brightness representative
of a t r u e out-the-window scene. With the high light flux, the film would be destroyed
almost immediately by infrared and ultraviolet radiation. To limit this damage, heat
absorption and rejection filters were used between the arc lamp and the film. Addi-
tional f i l m cooling was achieved by a very high, continuous airflow. This high airflow,
in turn, caused flutter and eventual destruction of the film. In the final mode of opera-
tion, a compromise was selected that accepted some film deterioration caused by the
heating and flutter and a scene illumination that required a relatively d a r k crew station.
In the simulator, the crew made use of the panel lighting and the s m a l l floodlights,
whereas in the spacecraft the sun illumination through the windows was generally
sufficient.
F o r other deficiencies in the system, practical solutions were never found, and,
as a result, procedural workarounds were developed to alleviate the impact on training.
One such deficiency involved the resolution of the imagery that was presented t o the
crewmember in the crew station. Although the inherent resolution of the direct-view
film system satisfied the specified requirements, copies of the film could not be prol-
duced consistently with the same high image quality. The processing of 100-foot lengths
of film does not produce results of the quality that can be achieved by copying individual
photographs. A 30-percent degradation was not uncommon. Another problem involved
the impossibility of maintaining the proper film positioning accuracy over the 80 feet of
film in each of the cassettes. Minor film stretching and warpage introduced e r r o r s of
significant magnitude. In addition, the drive system of the film did not have nearly as
tight a tolerance as required. Finally, a problem existed with the colors in the film.
While the films themselves contain very saturated colors, the projector tended t o mute
and mutilate the colors as displayed in windows - f o r instance, the blue oceans
appeared almost black in the windows.
In both the CMS and the LMS, a defect existed in that the star simulation would
bleed through the near-body scenes. The equipment, as originally designed, contained
hardware to occult the stars in a circular pattern of the s a m e diameter as the n e a r body
would subtend. This hardware proved difficult t o maintain, was nonlinear with position
and shape of the body in the FOV, and, as a result, did not successfully occult all the
stars. Finally, a shortcoming of all near-body simulations was the low scene bright-
ness. This condition required the crewmen t o d a r k adapt t o discern the necessary de-
tail out the window.
Significant funds have been spent in developing the near-body simulations f o r the
various space-flight vehicles. It is recommended that, in the future, the requirements
for out-the-window scenes be very carefully analyzed and that extreme c a r e be taken
to limit the simulator requirements to those obtainable with reasonably simple hard-
ware. For example, in the CMS, only the unity-power telescope in reality required
stringent tracking accuracies. The other four active windows could have been animated
with much less sophisticated hardware. Experience has also shown that window-scene
generators should be more integrated as in the LMS and the CMPS than the system-per-
window such as in the CMS. Such design considerations in future applications should
both reduce initial costs and result in lower maintenance effort and costs.
26
Landing Scenes
Landing scene simulation was required only in the L M S for training in the lunar
approach and landing phases. The altitude range desired was from a high-gate o r
breakout altitude down to and including touchdown. The scene was required t o provide
the crewmember with attitude and altitude information plus pertinent surface-feature
details. In all other programs (for instance, during earth entry and landing), there was
a minimum of c r e w activity relative to the window displays; therefore, little or no ex-
ternal scene simulation was supplied. In the LMS, the requirement w a s defined as a
continuous scene from an altitude of approximately 1 5 000 feet to touchdown. The orig-
inal requirement at the time of simulator procurement was for a generalized lunar s u r -
face containing representative features of the moon. This requirement was expanded
subsequently t o include modeling of the actual landing sites. Three-dimensional s u r -
face detail was required when surface irregularities became visible to the crewman.
For the lunar landing simulation, the attitude at which surface features became impor-
tant w a s defined as 2000 feet. An additional requirement was the casting of shadows
such as would be caused by local surface irregularities when illuminated with collimated
sunlight.
Overall, this system was determined to be inadequate for training. The major short-
coming was the lack of continuity from breakout to touchdown caused by the film system
problems mentioned in the discussion of near-body scenes. An additional shortcoming
was the lack of realism in the landing area model. An upgraded system, which used a
3-D model f r o m breakout altitude to touchdown, was developed for training of the first
lunar landing mission crews. Maximum altitude of the model simulation w a s increased
to approximately 12 000 feet, and the model scale was established at 1:2000. Two sig-
nificant features were added - an accurate model of the targeted landing area and a
collimated light source to illuminate the surface and produce lunar-terrain shadows.
Figure 13 is a photograph of the upgraded system. The model was viewed by an
27
Figure 13. - Lunar-surface model (scale 1:2000) of lunar module simulator.
articulated probe with a 110" FOV. To produce the large depth of field required f o r
viewing the surface obliquely, the probe had remote focus capabilities and tilt c o r r e c -
tion. The rotations in yaw and r o l l were unlimited, but pitch rotation was limited to
-10" to +110". The probe was driven horizontally a c r o s s the surface in latitude and
longitude, while the surface moved vertically to simulate altitude. The landing area
simulated was approximately 5 by 8 nautical miles. The image was relayed from the
probe to one window of the LM display system by the high-resolution closed-circuit TV.
Switching was available to supply the image to either of the two LM front windows.
Several undesirable characteristics still existed with the upgraded L&A system.
The closed-circuit TV performed according to expectations only with continuous main-
tenance. The horizon simulation, which was generated by mechanical rings around the
model, exhibited e r r o r s of 3" to 5 " f o r off-nominal trajectories. Duplication of the re-
flectivity of the lunar environment required continuous testing with each new surface
installed. As the desired sun elevation angle increased above 12", the probe shadow
became very prominent and the actual touchdown a r e a was in the shadow at angles above
18". Various means of floodlighting this area alleviated the shadow but compromised
the sun collimation.
28
Landing simulators used throughout the aircraft and airline industries for training
pilots are in some ways similar t o the L&A but are much less massive and, therefore,
have lower demands upon servosystems. New designs f o r spacecraft landing simulators
should take advantage of the design of these commercial systems. The use of colli-
mated light to illuminate model surfaces should be carefully analyzed. Continued ef-
f o r t s should be expended toward producing a brighter image on the TV screen. Tests
have been made with some degree of success on the use of black phosphors mixed with
bright phosphors t o provide a greater contrast ratio from less scattering of the light.
Other techniques should be studied. Although simulation of the lunar landing did not
require color (since the black-and-white system w a s very suitable), earth landing sys-
t e m s will probably require color; and the development of a tricolor high-resolution TV
system is desirable, particularly f o r night approach simulation.
Concluding Remarks
Beginning with a simple simulation requirement to provide an out-the-window
scene as a backup spacecraft-attitude reference in the Mercury flights, visual display
simulation s y s t e m s f o r the lunar landing mission have grown t o become a major but
necessary element of the Apollo mission simulators. For each visual task of the lunar
landing mission, a visual scene was simulated through an infinity optics system in the
mission simulator. Although the complexity and cost of these devices were significant,
their use in mission training was mandatory.
MOVING-BASE SIMULATIONS
Various moving-base Simulators have been used throughout manned space -flight
programs for crew training. The primary objective has been to familiarize flight c r e w s
with the dynamic environment and t o present them with visual realism of the operational
task. The use of moving-base simulators and the experience gained while developing
and operating such systems are described in this section.
29
simulator (fig. 14), the mobile partial-gravity simulator (fig. 15), the lunar landing
r e s e a r c h facility (fig. IS), and the LLTV (fig. 5).
Although training for tasks such as reentry, docking, launch abort, and lunar
landing has been accomplished on fixed-base simulators, crew proficiency f o r each
of these tasks has been improved with moving-base simulators. The moving-base de-
vices were used to add realism to the simulation by presenting realistic cues inherent
in body acceleration and binocular vision.
30
Application of Moving-Base Si mulators
Several moving-base simulators have
been used f o r crew training in environmental
familiarization and development of task per-
fection. In preparation for the early space
flights of Project Mercury, when little o r
no experience existed in the new environ-
ments, several devices were used t o famil-
iarize the crewmembers with anticipated o r
possible dynamic situations. The multiple-
axis spin test inertial facility (fig. 17),
which was a simulator built by the NASA
Lewis Research Center, consisted of three
gimbals individually powered by compressed-
gas thrusters. Angular rates from near
z e r o to more than 60 rpm could be gener-
ated. The crewmembers were whirled to
rotational speeds at which disorientation oc -
Figure 1 6 . - Lunar landing research curred t o familiarize them with this dis-
facility (Langley Research Center). orientation and t o enable them t o practice
stopping the motions by use of the Mercury
spacecraft hand controller and angular rate
instrument displays. The t e s t s served mainly t o build confidence that, in the-event of
a g r o s s automatic control system failure, the crewmembers could regain control of the
vehicle and stop all tumbling.
The human centrifuge (fig. 18) at the Naval Air Development Center, Johnsville,
Pennsylvania, was used f o r crew training in acceleration profiles and f o r engineering
evaluation of crew-station equipment. The gondola of the centrifuge was equipped with
the vehicle hardware required to support
the astronaut in launch and reentry phases
Figure 17. - Multiple-axis spin test Figure 18. - Centrifuge (Naval Air
inertial facility (Lewis Research Development Center, Johnsville,
Center). Pennsylvania).
31
of the mission. To further duplicate flight conditions, the gondola was evacuated to the
reduced cabin pressure of actual flight. Six-degree-of-freedom equations of motion
were used to animate the display panel and t o generate the acceleration profiles. In
the final training sessions for Project Mercury (August to September 196l), the crews,
equipment, and procedures were evaluated by simulation of the complete three-orbit
mission from preflight physical examination through countdown, launch, three orbits,
reentry, recovery, and postflight debriefings and physical examinations. Baseline
medical data for comparison with the flight data were derived from these sessions.
The ALFA t r a i n e r (fig. 12) was basically a pseudo-Mercury spacecraft mounted on a
spherical a i r bearing. This n e a r frictionless support, coupled with a combined
astronaut-plus-trainer center of gravity at the center of the ball, produced an accurate
simulation of spacecraft motion as far as attitude control was concerned. The struc-
t u r a l support systems and the effects of the one-g field on the t r a i n e r limited the yaw
and pitch motions t o *35 O. Roll was unlimited. All three attitude-reference systems
(periscope, panel instruments, and window) were available t o the crewmember. F o r
the periscope display, a 10-foot-diameter s c r e e n was viewed through optics simulating
the wide-angle view of the vehicle periscope. A moving earth scene of the orbital t r a c k
was back projected onto the screen. This display was relatively accurate f o r deviation
angles up to 25". For the panel displays, actual vehicle hardware was used to measure
the attitudes and rates and to display these values to the crewmember. F o r the view
through the window, a lighted horizon and a generalized s t a r field were provided. Mo-
tion of the trainer was controlled by the crewmembers' use of compressed-air jets.
Mercury spacecraft control systems were simulated. Disturbances that could be pro-
duced by misalinement of the retrorockets also were simulated by compressed-air jets.
The MSC translation and docking simulator (figs. 20 and 8) was a six-degree-of-
freedom simulator used primarily f o r crew training. The hardware consisted of a crew
station mounted in a three-axis gimbal. The gimbal was supported by an air-bearing
carriage to permit lateral motion. The two remaining translations, longitudinal and
vertical, were simulated by target motion. These two motions were reversed t o give
the proper directional illusion of crew-station motion. P r i m a r y emphasis was placed
on vehicle configuration and visual environment. Various control system modes were
32
Figure 20. - Translation and docking
simulator (Gemini).
The lunar landing r e s e a r c h facility (fig. 16) was used for r e s e a r c h and f o r crew
training associated with the piloting of a vehicle in simulated lunar gravity. This out-
side facility consisted of a gimbal-mounted vehicle suspended from a traveling bridge
crane. The piloted vehicle, originally with the crewman seated and l a t e r modified to
a standup configuration, had an operating envelope of 180 feet vertically, 360 feet lon-
gitudinally, and 42 feet laterally. Two operational modes w e r e used. The first mode,
used primarily for pilot indoctrination and s y s t e m test, simulated the lunar-gravity en-
vironment and rocket-engine lift through the use of the translation drive systems. The
second (simulation) mode used the main rockets and onboard f u e l to develop lift while
the translational drives supported five-sixths of the vehicle weight to simulate lunar
33
gravity. The simulation mode allowed ap-
proximately 2 minutes of operation whereas
the first mode afforded more than 20 min-
utes of continuous operation. Evaluations
of control system p a r a m e t e r s and piloting
techniques associated with lunar landing
touchdown were conducted. Operation with
various landing models allowed evaluation
in the area of obstacle avoidance and a
limited look at surface dust effects.
Simulation Experience
In general, the design, operation, and
modification requirements of moving-base
simulators are similar t o those of other
simulation hardware. However, in addition
to the g r e a t e r emphasis on crew safety re-
quired, some unique features must be con-
sidered in the design and operation of
moving-base simulators.
Figure 22. - Lunar landing training
vehicle simulator. Moving-base simulators usually are
designed f o r a special task o r s e t of tasks.
34
The design must concentrate on the simulation of those t a s k s while minimizing the un-
desirable side effects that are inherent in moving hardware. To overcome undesirable
side effects and to accomplish successful moving-base simulation, innovation and
unique approaches a r e needed, such as the separation of simulated acceleration cues
(seat of the pants) and velocity cues (visual), the use of airbearings to minimize other
motion cues such as rumble, special servosystem design to keep undesirable dynamic
cues outside the frequency and amplitude domain of sensory perception, rotation about
the individual body center of gravity to equalize force on extremeties, painting, lighting
to hide support or enclosure structure, and so on. The hardware design as well as the
computer software must take into account the man-in-the-loop aspect. Convenient in-
g r e s s and e g r e s s provisions are required and actual man-machine interface is neces-
s a r y ; however, only the pertinent interface should be included. Special programing,
such as drive-system compensation, may be required for proper hardware perform-
ance. An overall systems analysis should be performed before design finalization to
ensure that optimum simulation can be performed. Size and weight are basic consid-
erations in the design. The simulator size is dictated by flight-vehicle configuration,
and the design weight affects drive-system design and overall system complexity.
Structural integrity must be consistent with the dynamic environment and the man-
machine interface. Fidelity of the simulations is dependent on a satisfactory balance
of these parameters. Controls and displays usually are limited t o those necessary to
accomplish the unique task o r tasks. Cockpit displays other than those of the space-
craft are necessary in some moving-base simulators, such as the LLTV, because of
their operating requirements. Special effects of lighting and sound have been used t o
add significant cues for the performance of a given task. Lighting also can be used to
enhance the simulation. For example, in the MSC TDS, the target vehicle was illumi-
nated t o make it stand out against a subdued background. This technique is used to re-
duce the false cues given by supporting structures and surrounding walls. Foremost
in the mind of the designer must be the characteristics of the spacecraft or the situa-
tion being simulated. If characteristics peculiar to the simulator overshadow the situa-
tion o r spacecraft being simulated, the simulator will be of little value.
35
area is unique because a limited number of simulators of a particular configuration are
built, possibly only one. Therefore, a preliminary fabrication and checkout period, in
which prototype hardware is developed and tested before production, could be used to
develop better hardware and to eliminate operational problems. Another design consid-
eration should be the use of off-the-shelf i t e m s t o minimize special equipment, which
is hard to replace and usually expensive. In addition, it should be remembered that
simulation equipment has a high use rate and needs to be designed with a high degree of
reliability. The computer program (analog o r digital) will in some ways be unique be-
cause of the interface requirements with moving hardware. The nature of the hardware
must be considered and special formatting o r signal conditioning accomplished to per-
f o r m a valid simulation. In addition, it is useful to be able to verify the computer pro-
gram independent of the moving hardware. A computer self test, o r program designed
t o close the loop on itself, can be a valuable tool in program verification and checkout.
Concluding Remarks
Various moving-base simulators have been used in manned space flight to add a
degree of fidelity in certain flight regimes not available in fixed-based simulations. In
spite of design complexities, additional operational procedures, and safety considera-
tions (compared t o fixed-base devices), moving-base simulators will continue to play a
significant role in training and simulation f o r future programs.
36
(PDR) and formed the basic mathematical model and hardware design definition f o r each
respective simulator. At the PDR, the simulator reached a contract definition stage
(usually, although somewhat erroneously, referred to as a contractual "design freeze"),
beyond which all simulator updates o r changes to the spacecraft baseline data package
constituted a design change and required a corresponding simulator contractual change.
It is at this point that the t e r m s "modification, '' "configuration control, '' "in-line modi-
fications, '' and "modification kits" assume their f u l l significance and the execution phase
of modification activity begins. This modification activity usually commences prior to
f o r m a l contract delivery of the basic simulator to NASA.
1. Monitor and review regularly the agenda and minutes of the spacecraft Con-
figuration Control Panels (CCP's). This activity will normally cover all proposed and
approved requests for engineering change proposals (RECP's) from NASA and all en-
gineering change proposals (ECP's) from the contractor of less than a certain dollar
value o r with less than a certain schedule impact.
3. Obtain copies of, and review, the prime contractors' internal change authori-
zation paperwork. In the Apollo Program, these authorization documents were the
manufacturing change request (MCR) and the lunar module change request (LCR) from
the prime contractors f o r the CSM and the LM, respectively. These documents are
normally released to initiate a design study and frequently a r e weeks ahead of RECP,
ECP, o r CCA action. Consequently, the use of this monitoring tool allows routine
identification of proposed changes much earlier than the activities described previously
(items 1 and 2); however, c a r e must be exercised in cross-checking these releases
against the NASA CCP and CCB activity f o r approval authorization. An alternative to
this procedure is t o obtain the contractor's MCR release-to-manufacturing milestone,
37
which signifies that the MCR is approved f o r implementation. It should be noted that
there is no obligation for the contractor always t o release an MCR (or a s i m i l a r type
paper) for a change, and frequently the contractor does not do so f o r class I1 changes.
As mentioned earlier, certain shortcuts can be taken t o decrease the NASA man-
hours required to review and control the r a t h e r large bulk of information and documen-
tation identified previously. Examples of the shortcuts that have been used on the
Apollo Program and that should be considered on future programs are as follows:
1. Have the spacecraft contractor categorize MCR's and filter out those clearly
not applicable to simulator design. Examples are ground-support equipment MCR's that
are not transmitted to NASA simulation personnel.
2. Do not submit any simulator change for approval unless the spacecraft counter-
part is released t o manufacturing o r the item is considered to be a long-lead-time type
change. In the past, the released-to-manufacturing milestone has been identified on the
MCR document; if it is not, a tabulation of this information is required from the
contractor.
38
procedure is not mandatory. However, these persons usually possess the firsthand ex-
perience and knowledge needed to recognize candidate simulator improvements. When
simulator history is traced f r o m the first Mercury procedures t r a i n e r to the CMS and
the LMS, a dramatic increase in simulator improvement modifications is apparent.
This increase can be attributed, in large part, to the justifiable increase i n complexity
in the man-machine interface. F o r example, whereas the Mercury procedures trainer
was a rather simple procedural analog training device designed t o simulate a somewhat
repetitive series of missions, the CMS and LMS, each with a powerful computing com-
plex and complicated MCC interface, were designed to simulate a much l a r g e r reper-
toire of missions.
The process of identifying and obtaining the data listed in i t e m s 1 to 3 is simple, but a
difficulty inherent in item 4 is the undesirability of receiving updates to 500 000 space-
craft drawings. The method that works w e l l on the Apollo Program is correlation of
the spacecraft prime contractor MCR or LCR to newly defined spacecraft drawings or
engineering erder (EO) to existing drawings, m d transmittal to NASA of the drawings
o r EO'S on microfilm aperture cards. Furthermore, NASA reviews MCR and LCR
documents immediately after they are released and informs the contractor which poten-
tial changes will have impact on the simulator; consequently, particular data are iden-
tified long before the data are released, which is a desirable condition. Also, this
procedure eliminates transmittal of data for MCR's that do not have impact on a simu-
lator. The document that relates MCR's to drawings and EO-type data is called a data
submittal letter and is a cross-tabulation of MCR number to drawing or EO number.
After an MCR o r LCR is defined as applicable to a simulator, all spacecraft data re-
leased against it will be supplied to NASA.
All proposed modifications for the simulator submitted to the Simulator Control
Panel (SCP) require preparation of a standard form known as a modification request
(MR). The MR is required to contain the following information:
1. The change t o the actual spacecraft and its effect on vehicle operation o r crew-
member interface (or both) (not required for simulator improvement modification)
39
3. The effect on crew training if the change is not implemented
6. An estimate of parts cost and source, and the man-hours and computer-hours
required to implement the modification f r o m preliminary design to final NASA
acceptance
2. Evaluate all other proposed changes, such as launch vehicle, mission trajec-
tory (reset), and simulator improvement, that have impact on the simulator.
b. Crew-training requirements
c. Simulator availability
f. cost
40
8. Act as the overall configuration management control point by defining any addi-
tional procedures required for a specific modification, assigning action items, and en-
suring that the overall modification cycle is operating in a smooth and timely manner.
2. Determine that the M R is not yet impacted and that impact is required for con-
sideration at the next SCP meeting.
3. Determine that the MR is not impacted because of lack of data and that action
is required to obtain data.
7. Determine that the MR revision has no effect, and, thus, that the previous re-
vision level remains in effect.
1 Modification ’ 3
9. Determine that the MR is approved
requirement Pre-SCP
determination review for implementation, but that new impact on
a.ndimpaci GFE o r CFE hardware is required f o r the
Modification Modification
SCP Approved -’preliminary
. final design from the appropriate MSC directorate, from
review design and * and NASA the NASA simulator support contractor
NASA approval approval
Modification
installation
checkout and
NASA acceptance
Modification- ’
documentation
(nonconcurrent)
acceptance
- updateand
acceptance
discussed. The SCP receives and reviews
those modifications recommended for ap-
proval or disapproval at the pre-SCP meet-
41
The NASA simulator support contractor SCP responsibilities and participation
a r e as follows:
2. Assure that the necessary spacecraft data are collected and reviewed to the
extent necessary for impact on the modification for the SCP. This impact entails the
following activities:
e. Estimate any required parts cost and indicate the parts sources.
5. Understand and implement the ECP decisions and assigned action items with
proper NASA coordination.
42
Configuration Accountability
Configuration accountability involves the establishment, monitoring, and updating
of simulator documentation to define a baseline and a modified current configuration of
a simulator at any point in time in t e r m s of hardware and software. This process re-
quires the following:
4. Standardization of items 1 t o 3
All simulator hardware has required and will continue t o require comprehensive
documentation. This documentation ranges from the basic drawing tree definition to the
specific drawings defining fabricated and purchased hardware as well as locating the
hardware within the simulator complex. However, within this large m a s s of documen-
tation, the basic requirements and procedures for updating, providing traceability, and,
in general, accountability are the s a m e for every hardware document i n the system.
Thus, f o r purposes of this paper, discussion will be provided for a single piece of hard-
ware docume-ntation entitled a drawing, with the realization that, although this drawing
may have dozens of variations and specific uses, for purposes of configuration manage-
ment it is one homogeneous entity.
The following facts have been proved true for hardware accountability:
2. One piece of paper must be generated for every simulator hardware change.
(This paper has had and may use any one of a number of names, but it is most univer-
sally known as an EO.)
3. An EO must describe before and after conditions when the EO makes a hard-
ware change in an existing drawing.
4. New drawings must be authorized by EO'S. Tne EO'S place the new draWiiiS
in their respective positions within the drawing tree by making proper references t o
the next higher and lower drawings in the tree.
43
6. The quality control section performs the final acceptance inspection of an
EO by comparing it to the affected, parent, and updated drawings and t o the hardware.
7. The EO's may be allowed to build up against a drawing but never t o the extent
that the usefulness of the drawing is destroyed.
10. The exact hardware configuration of any simulator can be tracked and iden-
tified by tracking all EO's with the baseline drawings (item 1).
44
fail-safe capability that prevents the affected program or subroutine from degenerating
t o a state any less desirable than that prevailing before a modification.
7. All approved modification and discrepancy patches (SCR's) are recorded and
approved for the next scheduled assembly. It is highly desirable to minimize this re-
cording effort by automating (computerizing) the off-line capability to do so. (At the
time of this writing, this activity is in work for the Apollo Program simulators. 1
10. When a software mathematical model is called for, all open approved SCR's
a r e provided automatically.
12. The octal training tape technique also should be examined f o r usefulness and
application to a specific simulator program. This approach uses two octal tapes, one
containing all newly created work and the other containing all newly approved work.
Work is shifted (approved) from one tape to the other and, f o r crew training, the load
2nd the octal tape with approved softx72re are used. Naturally, the next new assembly
adds to this approved software and the cycle repeats itself. The advantage of this oper-
ation is that new software work can be used a l m o s t immediately without waiting for a
new assembly. At the same time, approved software is segregated from developmental
o r unapproved software and preserves the configuration and operational integrity of the
simulator.
Concluding Remarks
Because of the continuing changes of both the spacecraft and mission, and the
necessity for a correspondingly prompt and accurate method of updating the flight crew
simulators, a systematic method of simulator configuration management is a confirmed
requirement. The Apollo Program provided the most s e v e r e requirements and prob-
lems encountered t o date in this area. These requirements and problems were satis-
factorily met and resolved. The viable methods of simulator configuration management
that were developed and refined during the Apollo Program should lend themselves di-
rectly to future programs.
45
CONCLUSIONS
REFERENCE
1. La France, Robert C. : Sixteen Earth-Orbit Film for the Apollo Mission Simulator.
Information Display, vol. 6, May-June 1969, pp. 65-70.
,
46
A PPEND IX
SIMULATOR DESCRIPTION
PROJECT MERCURY
Centrifuge
The centrifuge located at the Naval Air Development Center, Johnsville, Pennsyl-
vania (fig. 18), was used to provide the Project Mercury crews with training f o r the
launch, launch abort, and entry phases under simulated acceleration loads. The accel-
eration simulation was mechanized in both closed- and open-loop modes through the
Aviation Computer Laboratory analog computers. Initial runs were made in the closed-
loop mode, in which the resulting acceleration profile as well as the panel displays
responded t o inputs from the pilot's hand controller. Analyses of these first runs indi-
cated that adequate training could be obtained by simply preprograming the acceleration
profiles (open loop). In subsequent centrifuge training, only the displays were integra-
ted with the computer to respond to the pilot's inputs. The cockpit equipment included
individual crew couches, restraint systems, the spacecraft hand controller, and a por-
tion of the spacecraft instrument panel with emphasis on the sequence telelights.
Training was accomplished with both soft and hard pressure-suit operations. Manual
control of the spacecraft, including manual retrofire maneuvers and simulated system
failures during launch and entry, was covered in the training.
47
reaction-control forces. For the cockpit displays, actual spacecraft hardware was
used to measure the attitudes and rates and to display these values to the pilot. A sim-
ulated out-the-window display was provided. (A detailed description is provided in the
section entitled "Moving-Base Simulations. 'I)
Part-Task S i mulator s
Two part-task simulators (fig. 24) provided the first simulation of the astronaut's
control tasks in Project Mercury. Both simulators used analog computers. The tasks
simulated included changing spacecraft attitude in orbital flight (including orientation
for retrofire), holding r e t r o f i r e attitude in
the presence of retrorocket misalinements,
and damping pitch and yaw oscillations
through the entry.
GEMINI PROGRAM
48
Translation and Docking Simulator
The TDS (fig. 20) was designed t o provide crew training for the last 100 feet of the
orbital rendezvous - the final closure and docking. The simulator provided six de-
g r e e s of freedom with the proper combination of motion of the crew station (spacecraft
mockup) and the simulated Agena target vehicle. Both the spacecraft carriage and the
target vehicle were supported on air bearings to provide the low friction required to
simulate the small, precise motion of orbital flight. Various lighting conditions and
control modes were provided in the training operation. In addition to the docking ma-
neuver, the simulator provided a training capability for formation flying and target-
vehicle inspection. The simulator was driven by analog computers.
Centrifuge
The Naval Air Development Center centrifuge (fig. 18) at Johnsville, Pennsyl-
vania, was used to provide launch, launch abort, and entry acceleration training for the
Gemini astronauts. The cockpit was configured to simulate the Gemini command pilot's
station. The acceleration profiles were preprogramed; however, the entry dynamics
as displayed to the pilot on his cockpit instruments were closed loop with the analog
computer. Suited runs were made to include the effect on the control task of both soft
and hard (pressurized) modes of suit operation.
APOLLO PROGRAM
49
guidance computer. A significant portion of the LMS, which provided training for final
touchdown, was the L&A visual display system. This system included a lunar t e r r a i n
model (scale 1:2000) of the specific landing site f o r each mission.
50
Centrifuge
The MSC centrifuge (fig. 25) was
used to provide Apollo Program crew train-
ing in the u s e of the entry monitor system
during simulated lunar-return entry and
associated acceleration conditions. The
simulated cockpit in the gondola included
three crew couches, an entry monitor sys-
tem, a flight director attitude indicator, a
gravity meter, and the right-hand rotational
hand controller. Five computers, three
analog and two digital, w e r e used to simu-
late the vehicle dynamics and t o drive the Figure 25. - Centrifuge (Manned Space-
centrifuge. craft Center).
51
Partial -Gravity Si mulators
Two dynamic simulators were designed and operated to provide astronaut training
for the 1/6-g lunar-surface operations. Both simulators incorporated a uniquely de-
signed overhead support system, which provided lifting f o r c e equal t o five-sixths of the
astronaut's total weight (suit, backpack, etc. ). The vertical servosystem of the partial-
gravity simulator (fig. 14) was suspended f r o m an overhead monorail. This concept
was subsequently adapted to a truck body f o r the mobile partial-gravity simulator
(fig. 15) and thereby provided essentially unlimited fore-and-aft mobility.